Spectrochimica Acta Part B 70 (2012) 45–50 Contents lists available at SciVerse ScienceDirect Spectrochimica Acta Part B journal homepage: www.elsevier.com/locate/sab Quantitative laser-induced breakdown spectroscopy of potassium for in-situ geochronology on Mars Christopher B. Stipe a,⁎, Edward Guevara a, Jonathan Brown a, George R. Rossman b a Department of Mechanical Engineering, Seattle University, Seattle, WA 98122, USA b Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA 91125, USA article info abstract Article history: Laser-induced breakdown spectroscopy is explored for the development of an in-situ K–Ar geochronology instru- Received 30 December 2011 ment for Mars. Potassium concentrations in standard basaltic glasses and equivalent rock samples in their natural Accepted 24 April 2012 form are quantified using the potassium doublet at 766.49 and 769.90 nm. Measurement precision varies from 0.5 Available online 3 May 2012 to 5.5 (% RSD) over the 3.63% to 0.025% potassium by weight for the standard samples, and little additional preci- sion is achieved above 20 laser shots at 5 locations. For the glass standards, the quantification limits are 920 and Keywords: 66 ppm for non-weighted and weighted calibration methods, respectively. For the basaltic rocks, the quantification Laser-induced breakdown spectroscopy Geochronology limits are 2650 and 328 ppm for the non-weighted and weighted calibration methods, respectively. The heteroge- K–Ar dating neity of the rock samples leads to larger variations in potassium signal; however, normalizing the potassium peak Potassium by base area at 25 locations on the rock improved calibration accuracy. Including only errors in LIBS measurements, Mars estimated age errors for the glasses range from approximately ±30 Ma for 3000 Ma samples to±2 Ma for 100 Ma samples. For the basaltic rocks, the age errors are approximately ±120 Ma for 3000 Ma samples and ±8 Ma for 100 Ma samples. © 2012 Elsevier B.V. All rights reserved. 1. Introduction for measure by an array of techniques in multiple laboratories, the engineering challenges associate with a robotic launch from the surface The study of Martian geological history is hindered by the absence of Mars are substantial. Maybe more importantly, the returned sample of direct absolute age characterization of surface materials of known approach does not allow for iteration between sample selection and geographical location. While the relative chronology of Martian surfaces corresponding age measurements; thus, if the correct compositional has been accomplished through stratigraphic studies [1],adirect or age range of rocks is not collected to accurately correlate crater measure of absolute age has not yet been achieved. Instead, indirect densities, a costly second mission would be required. In-situ measure- estimates of absolute age are extrapolated from lunar cratering densities ments would allow for this experimental iteration, at the cost of precision correlated to age-dated return samples from Apollo and Luna missions and accuracy. The best approach may be to include an in-situ technique [2,3]. Known sources of errors for this approach include secondary in a sample return mission as a sample selection tool. Potassium–Argon craters formed by impact ejecta and crater resurfacing by erosion dating shows promise as an in-situ technique suitable for Mars [8]. and deposition [4,5]. Hartmann estimates a factor of 2–4 uncertainty As envisioned, a possible system will consist of two instruments: in age dating when extrapolating from lunar to Martian cratering (1) a heating chamber or laser heating system with a mass spectrometer densities [5]. Additionally, it has been shown that differences in crater to measure outgassed argon and (2) a laser-induced breakdown spectros- counting techniques lead to age determinations that vary between 5% copy (LIBS) instrument to measure potassium. In our paper, we evaluate and 20% [6]. Precisely aged Martian meteorites shed light on the planet's LIBS for quantifying potassium abundance. volcanic activity and the existence of old geological material (ALH84001) LIBS is a spectroscopic method where atomic emission from a laser [7]; yet, they cannot be used to correlate cratering densities because the generated plasma is collected to identify and quantify elemental place of origin of these meteorites is unknown. abundances [9]. The Mars Science Laboratory rover, launched in the Both sample return and in-situ geochronology measurements are fall of 2011, includes a LIBS instrument as part of the ChemCAM being explored to solve this problem, each with associated advantages suite [10,11]. The performance of LIBS for elemental quantification of and disadvantages. While returned samples provide the best option for geological materials has been extensively studied under simulated obtaining accurate age measurements, as the samples will be available Martian environmental conditions. Knight et al. observed signal enhance- ments of 3–4 times under Martian atmospheric pressure and composition compared to the atmospheric conditions on earth [12]. This signal ⁎ Corresponding author at: 901 12th Avenue, Seattle, WA 98122. Tel.: +1 206 296 6941; fax: +1 206 296 2209. enhancement is due to the reduced pressure which improves laser E-mail address: [email protected] (C.B. Stipe). ablation of material from the surface by reducing attenuation of the 0584-8547/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.sab.2012.04.010 46 C.B. Stipe et al. / Spectrochimica Acta Part B 70 (2012) 45–50 laser energy by plasma shielding [9]. Clegg et al. have explored multi- with correcting for excess 40Ar will be addressed in a later publication variate spectral analysis methods to reduce matrix effects [11], and ac- describing the mass spectrometer measurements. curacies on the order of ±10% have been achieved for most elements In this study, we investigate the efficacy of LIBS for quantifying po- [13]. Dell'Aglio et al. performed LIBS measurements on six Martian, tassium in standard basaltic glass and natural basaltic rock samples. lunar, and other meteorites [14] to identify and quantify elements for Experimental parameters are explored to understand their influence on geological classification. For major elements with weight concentrations LIBS potassium signal. Potassium calibration curves are generated to de- above 6%, prediction errors ranged from approximately 0.5% to 15%. For termine measurement accuracy and resulting errors in age quantification. minor elements, prediction errors were as high as 60%. Thompson et al. analyzed two Martian basaltic shergottite meteorites (DaG 476 and 2. Experimental Zagami) by LIBS and achieved oxide weight percent differences from literature values from 4% to 79% [15]. Again, higher concentrated Laser-induced breakdown spectroscopy measurements were per- elements were better predicted than elements of lower concentration. formed with a Photon Machines Insight™ LIBS system. A representative Colao et al. studied the feasibility of LIBS as an in-situ instrument schematic of the experimental apparatus was published previously [19]. by measuring NIST standard reference soils and Martian rock ana- The laser-induced plasma was generated by a frequency quadrupled logues [16]. Quantifying potassium in NIST standard reference soils Nd:YAG laser emitting 266 nm, 7 ns laser pulses with a beam diameter pressed into pellets, they achieved accuracies of prediction between of 5 mm, a maximum energy of 35 mJ, and a shot-to-shot energy stabil- 12% and 30%. The predicted concentrations were consistently lower ity of ±2%. The 266 nm light was better absorbed by the basaltic glass than known concentrations, which the authors attribute to signal satura- samples than the fundamental and doubled laser frequencies. After tion and self-absorption. Measuring potassium in silicate materials, LIBS exiting the laser, the beam is expanded, collimated, and then focused predicted concentrations were approximately 40% lower than those onto the sample surface using a 5× microscope objective. The laser measured by EDX. In their study, eleven elements were measured simul- spot size at the sample surface is controlled by varying the distance taneously, making it difficult to optimize experimental parameters for between the beam expander and collimator. Imaging of the sample any one specific element. In our study, we focus solely on measuring po- surface with a video camera ensured a constant lens-to-sample distance tassium by LIBS in an effort to develop an in-situ K–Ar geochronology for samples of different thicknesses. Optical emission was collected instrument. through a 400 μmdiameterfiber optic cable aimed 50 degrees from the Currently, only a small amount of preliminary work exists in the vertical axis. The collected light entered a 50×50 μm slit into an echelle literature on the development of K–Ar dating using LIBS to quantify spectrometer (Catalina Scientific model SE 200) with a resolving power potassium. Swindle et al. [17] provide a LIBS calibration curve for of λ/Δλ~1700. The dispersed light was imaged by an iCCD camera samples containing up to about 6% potassium by weight. Solé [18] (Andor iStar model DH734i). Chip read-out time limited the acquisition also proposes the technique, showing a representative spectrum of rate to 1 Hz. potassium by LIBS. The three naturally occurring potassium isotopes of 39K(93.2851%), 3. Samples 40K(0.0117%),and41K (6.7302%) are simultaneously